DESCRIPTION: Pancreatic cancer, mostly adenocarcinoma of the exocrine pancreas, is the 5th leading cause of cancer deaths, totalling an estimated 26,300 new cases and 27,800 deaths in 1996. Survival is abysmal and no early screening methods are currently available. Suspected environmental determinants include tobacco use, various occupational exposures, high dietary fat intake, and perhaps ionizing radiation. Suspected genetic determinants include somatic mutations in the oncogene k-ras and various chromosome regions (1p, 3p, 6q, 8p, and 17p, seen cytogenetically; 5q, 8p, 1p33, and 11q13, seen by loss of genetic heterogeneity), and rare Mendelian conditions (e.g., hereditary pancreatitis, hereditary dysplastic nevus syndrome, hereditary nonpolyposis colorectal cancer, hereditary heart and ovarian cancer due to BRCA2, and the Li-Fraumeni cancer family syndrome). Family studies, using the proposed strategy have previously yielded insights into the origins of various cancers, including colorectal, breast, and sarcoma; the same approach is likely to be useful in pancreatic cancer. Specific Aim 1 is to update familial aggregations of adenocarcinoma of the pancreas. A national registry of families with pancreatic cancer has been initiated and the plans in this application are for updating the 71 current families and 13 provisional families, and adding new ones to a total 150 within 4 years. Environmental and medical questionnaires will be completed for pancreatic cancer patients and a control. Specific Aim 2 is to acquire biologic specimens from registered families and maintain a specimen repository. The history of all cancers will be documented with medical records; slides, tissue blocks, DNA and sera will be assembled on available patients and their first degree relatives. Specific Aim 3 is to conduct preliminary analyses and assays on registry material. Collaborations will be struck to conduct pilot work on the specimens (viz., for mutations in TP53, BRCA2, activated k-ras and other oncogenes, dinucleotide repeats instability, cytogenetics, and review of histopathology). Serial sera from first degree relatives, who can be considered at increased risk, will be assayed for a battery of tumor markers, including CEA, CA125, CA19-9, and CA242, to prospectively detect pre-symptomatic cancer.